A free Claude Code skill that extracts a structured brand voice profile from any website. One URL in, a portable voice document out.
Your website already has a voice. This skill quantifies it — 6 dimensions scored, vocabulary patterns mapped, positioning signals identified, and 15-20 starter guardrails generated. The output is yours to keep, share, and use anywhere.
| Step | What Happens |
|---|---|
| 1. Scrape | Pulls your homepage, about page, and blog to analyze voice patterns |
| 2. Dimension | Scores your voice across 6 measurable dimensions (each 1-5) |
| 3. Vocabulary | Extracts signature phrases, power words, and notably absent language |
| 4. Position | Maps what you call yourself, who you serve, and how you differentiate |
| 5. Guardrails | Generates 15-20 specific, enforceable content rules based on YOUR voice |
| 6. Export | Outputs a portable voice profile you can copy and use anywhere |
| Dimension | What It Measures |
|---|---|
| Authority | Tentative → Absolute. How much command does the voice carry? |
| Formality | Casual → Academic. How structured is the communication? |
| Emotional Temperature | Clinical → Hype. How much feeling is in the writing? |
| Specificity | Vague → Surgical. How concrete vs. abstract is the language? |
| Buyer Orientation | Self-centered → Buyer-obsessed. Who is the writing about? |
| Rhythm | Monotone → Crafted. What's the sentence cadence? |
One command:
git clone https://github.com/whystrohm/whystrohm-voice-extract.git ~/.claude/skills/whystrohm-voice-extractOr manually: Download this repo and copy the whystrohm-voice-extract/ folder to ~/.claude/skills/.
- Claude Code (CLI, desktop app, or IDE extension)
- No API keys. No external tools. No configuration.
/whystrohm-voice-extract
You'll be asked for one thing: your website URL. That's it.
The extraction takes about 60 seconds.
═══════════════════════════════════════════════════
VOICE PROFILE: yourcompany.com
═══════════════════════════════════════════════════
VOICE DIMENSIONS
────────────────────────────────────
Authority: 4/5 ████████░░
Formality: 2/5 ████░░░░░░
Emotional Temp: 3/5 ██████░░░░
Specificity: 3/5 ██████░░░░
Buyer Orientation: 4/5 ████████░░
Rhythm: 3/5 ██████░░░░
────────────────────────────────────
Followed by:
- Per-dimension evidence with exact quotes from your site
- Signature phrases, power words, and notably absent vocabulary
- A positioning summary (what a first-time visitor sees)
- 15-20 starter guardrails derived from YOUR voice patterns
- A portable profile you can copy and use immediately
This is a diagnostic tool. It tells you what your voice looks like, quantified. You can use the output to brief writers, evaluate content, or build internal documentation.
This is not a system. A profile describes your voice. A system enforces it. The difference: a profile gets ignored under deadline pressure. A system catches drift before it ships — automatically, every time.
whystrohm-voice-extract/
├── SKILL.md # Orchestrator — controls the extraction flow
├── rules/
│ ├── voice-dimensions.md # The 6-dimension scoring framework
│ ├── vocabulary-analysis.md # Vocabulary fingerprint extraction
│ └── guardrail-generator.md # How starter guardrails are derived
├── templates/
│ ├── voice-profile.md # Output formatting
│ └── cta.md # Closing pitch
├── .github/ # Issue templates, PR template
├── CONTRIBUTING.md
├── SECURITY.md
├── LICENSE # MIT
└── README.md
| Skill | What It Does | Install |
|---|---|---|
| Content Audit | Score content against 5 layers, get a live rewrite | git clone ...whystrohm-audit.git ~/.claude/skills/whystrohm-audit |
| Voice Scorer | Measure voice drift between your website and social content | git clone ...whystrohm-voice-scorer.git ~/.claude/skills/whystrohm-voice-scorer |
This skill extracts a voice profile in 60 seconds.
A full content infrastructure install takes that profile and turns it into 40-60 enforceable guardrails, a video production pipeline, multi-platform automation, and a content engine your team owns completely. Built in 30 days. You own everything. No retainer.
Score your content: whystrohm.com/scan Book a call: whystrohm.com/audit
WhyStrohm builds content infrastructure for founder-led companies doing $500K-$5M. One person. 10+ years in defense systems engineering. No templates. No outsourcing. Systems that run your brand — built and handed over.
MIT — see LICENSE.